Food cost

What your plates actually cost this week, against what you charged for them.

Food cost in SigmaPointPi
/verticals/restaurant/food-cost

Food cost percentage is theoretical cost against actual sales. Theoretical is what your recipes say the food should have cost for the covers you served. Actual is what you spent. The gap between them is waste, theft, over-portioning, or a supplier price that moved without telling you.

This page computes both from invoices and sales and shows the variance by category, so a jump in protein cost does not hide inside a stable overall percentage.

Weekly
The period that matters. Monthly food cost tells you about a problem after four weeks of it
Theoretical vs actual
Both computed, with the variance broken out by category rather than blended
Per invoice
Supplier price movements detected line by line as bills arrive

Where everything sits

Food Cost Analysis
Food Cost Analysis
Actual Food Cost
Actual Food Cost
Monthly Food Spend
Monthly Food Spend
Monthly Revenue
Monthly Revenue
Cost Variance
Cost Variance
No theoretical-vs-actual variance data yet.
No theoretical-vs-actual variance data yet.

How to work this page

Get recipes costed first

Theoretical cost is impossible without recipes. Cost the twenty items that make up most of your covers and you have most of the value, without costing the whole menu.

Read variance by category

Protein, produce, dairy, dry goods, beverage. A single overall percentage hides the category that moved.

Watch supplier price drift

The same item across successive invoices is compared. A price that rose quietly over six weeks shows as a line, not a surprise.

Act on the gap, not the percentage

A 31 percent food cost is fine or terrible depending on your model. The theoretical to actual gap is the number that means the same thing in every restaurant.

On a phone

Food cost on iPhone 15 Pro Max

Every figure from the desktop appears here, stacked rather than reduced. Tables scroll inside themselves so the page never moves sideways, and figures keep their separators and their alignment at every width.

Questions people actually ask

How does this handle waste?

Recorded waste reduces the gap because it explains it. Unrecorded waste sits inside the unexplained variance, which is the point of separating the two.

What about items sold below cost deliberately?

Menu engineering handles those. A loss leader is a decision, and it is marked as one so it does not read as a costing error here.